Integrating subgroups with mixed-type endpoints in early phase oncology trials

Area of Science:

• Oncology
• Biostatistics
• Clinical Trial Design

Background:

• Testing anti-cancer agents across multiple disease subtypes presents significant challenges, especially with mixed endpoint types like tumor response and progression-free survival.
• Current oncology practices often involve parallel, independent screening trials for patient subgroups, which may overlook similarities in treatment response between subpopulations.

Purpose of the Study:

• To develop a simplified statistical approach for jointly modeling patient subgroups with mixed-type endpoints in anti-cancer drug trials.
• To enhance the efficiency of treatment effect estimation by enabling 'borrowing strength' across relevant subgroups.

Main Methods:

• A novel joint modeling framework was developed to accommodate both binary (e.g., tumor response) and time-to-event (e.g., progression-free survival) endpoints simultaneously.
• The methodology allows for the integration of data from multiple, distinct patient subgroups within a single analytical model.

Main Results:

• The proposed joint modeling approach facilitates more efficient estimation of treatment effects compared to independent subgroup analyses.
• Demonstrates the utility of borrowing statistical strength across subgroups that exhibit similar responses to therapy, leading to improved precision.

Conclusions:

• The developed method offers a more efficient and robust strategy for analyzing anti-cancer agent efficacy across diverse patient populations with mixed endpoints.
• This approach addresses limitations of traditional independent subgroup trials by leveraging shared information for more reliable treatment effect assessment.